Smooth pursuit eye movements respond to moving objects in an attempt to generate smooth eye movements that keep the eyes pointed at the moving object. The neural substrate for pursuit is well known. Visual inputs arise in the retina and pass through the lateral geniculate nucleus to the cerebral cortex. They are transmitted into extrastriate visual areas that process moving images, and through the pons and cerebellum to the final brainstem motor pathways of the eye movement system. This proposal seeks to understand how learning occurs within the pursuit system. If a monkey tracks a target that moves at an initial speed for 100 ms and then steps to a higher or lower speed, then learning occurs in the initiation of pursuit. Learning is expressed in the earliest part of the response, well before the first visual feedback about the change in target velocity. This implies that learning occurs in the feed-forward pathways that transform visual inputs into motor commands. The present application has 4 specific aims that will elucidate the neural basis for pursuit learning. 1) Determine the representation of pursuit learning by following the responses of neurons during the acquisition of learning. This experiment will be done while recording in the arcuate frontal pursuit area (FPA), the middle temporal visual area (MT), and the oculomotor vermis of the cerebellum. 2) Investigate whether learning occurs downstream from the FPA by recording the effect of learning on the smooth eye movements evoked by stimulation of the FPA. 3) Search for pathways that cause learning by asking whether learning is induced under artificial conditions that pair target motion with stimulation of the FPA. This experiment may cause a persistent increase in eye acceleration if the stimulation-induced enhancement of the initiation of pursuit has access to the site(s) of learning. 4) Develop a method that will allow eye movements to be evoked by stimulation of MT during fixation, and determine how learning alters those eye movements. These 4 aims will reveal whether learning occurs in sensory or motor areas, and where learning is localized relative to the sensory-motor processing of pursuit. It will also ask whether learning is implemented by changing the gain of visual-motor transmission, or by alt of that process as probed by on-line gain of visual-motor transmission, or by altering the modulation of that process as probed by on-line gain control. These experiments will provide fundamental knowledge about the cortico-ponte-cerebellar circuits that are commonly involved in neurological disorders.